Log of the #fcrepo channel on chat.freenode.net

Using timezone: Eastern Standard Time
* dwilcox joins08:13
* ajs6f joins08:41
* kat3_drx joins08:50
* dhlamb joins08:53
* whikloj joins09:05
* osmandin joins09:16
* mikeAtUVa joins09:20
* mohamedar joins09:23
* bseeger joins09:53
* github-ff joins09:55
[fcrepo-transform] escowles pushed 2 new commits to master: https://git.io/vgXbG
fcrepo-transform/master 32c7836 Andrew Woods: Ensure transformed dates are serialized as formatted text...
fcrepo-transform/master 863592a Esmé Cowles: Merge pull request #7 from awoods/fcrepo-1877...
* github-ff leaves
* dwilcox leaves
* peichman joins10:01
* mikeAtUVa leaves
* travis-ci joins10:02
fcrepo4-exts/fcrepo-transform#28 (master - 863592a : Esmé Cowles): The build passed.
Change view : https://github.com/fcrepo4-exts/fcrepo-transform/compare/67db439d9410...863592a6896b
Build details : https://travis-ci.org/fcrepo4-exts/fcrepo-transform/builds/108800938
* travis-ci leaves
* dwilcox joins10:14
* mcritchlow joins10:39
* dwilcox leaves10:44
* peichman leaves11:00
* peichman1 joins
* jcoyne joins11:06
<barmintor>I am already hurting for coffee. brb11:07
* jgpawletko joins11:08
<barmintor>awoods: I could probably throw up a similar strawman for fixity, though afaik there aren't analogous specs to memento to refer to. I guess you could reuse the memento timemap language for check history, and use the premis vocab?11:11
<awoods>on a call
<barmintor>but I think there is virtue in trying to build on basic LDP mechanics, which would make things more easily back-filled.
* dwilcox joins11:32
* bseeger leaves12:04
<awoods>barmintor: a fixity strawman would be very welcomed... in line with LDP vs. shoe-horning into Memento12:05
<ajs6f>barmintor: +1 to that approach. PREMIS might have some terms for you.12:06
barmintor: What's more, we will need to factor out the means of extending LDP from between versioning and fixity.12:08
* peichman1 leaves12:10
<barmintor>ajs6f: you are underscoring my hesitation re: headers
* peichman joins12:11
<ajs6f>barmintor: We gots to do what we gots to do. Magic URIs are not better, right?
<barmintor>I think they are not, in part because headers encourage more adaptable clients12:12
<ajs6f>barmintor: Right, right. We only have a few tools to hand, and since it's REST, that's by design.
<barmintor>is premis:Fixity used as a containing element for premis:Events?
the ontology is a little opaque12:13
<ajs6f>barmintor: You know more than I. Alls I was saying was that people seem to find it useful. I think escowles has looked into it more deeply.
<escowles>barmintor: ajs6f: i can look (i need to read the new version of the data dictionary anyway)12:14
<barmintor>well LoC is down for maintenance, so use the googly cache12:15
escowles: http://webcache.googleusercontent.com/search?q=cache:1bdvNsB1aW8J:id.loc.gov/ontologies/premis.html+&cd=2&hl=en&ct=clnk&gl=us
<escowles>barmintor: hmm, http://www.loc.gov/standards/premis/v3/premis-3-0-final.pdf worked for me (maybe in my browser cache tho)
<barmintor>possible that the html is being served by something that is being maintained, but pdf is static?12:16
* barmintor shrugs
* jcoyne leaves
<barmintor>maybe I just had bad luck, or I'm on a LoC watchlist or something
<escowles>barmintor: there are worse do-not-fly lists to be on
* jcoyne joins12:19
<escowles>barmintor: there is a link between premis:Fixity and premis:Event (the action of computing the fixity is recorded as an event, with just the digest and algorithm being in premis:Fixity12:20
since we put the algo and digest in a urn, then i think we don't need any premis:Fixity records at all -- just the events recording the fixity checking actions
<ajs6f>escowles: In an LDP container?12:21
<barmintor>escowles: is there a class for an event stream?
<escowles>ajs6f: well, i think they could be audit events (in a container if you have the module), or just JMS events otherwise
<ajs6f>escowles: I suppose it depends on what people want to do them. Can we reuse the pattern you used for audit?12:22
<escowles>barmintor: just the regular premis:Event class -- i don't think there is anything in premis to aggregate them
<barmintor>escowles: I **think** it might be good to record the algo and value that was used to verify in the event? Would be more discrete, not require a bunch of memento-ing the LDP-NR to find out. Or maybe a client just doesn't care?12:23
<escowles>ajs6f: the complication is that all of the existing events are triggered by JCR events. ermadmix took a look at generating our own events, but IIRC, it was kind of a pain
<barmintor>fixity is for suckers.
* barmintor embraces change lives in the now
<escowles>barmintor: i bet some people care (e.g., if they do multiple algorithms), so putting that in the event record somewhere seems good12:24
<ajs6f>barmintor is a cosmopolitan.
<escowles>the premis ontology has a bunch of wrappers for different kinds of event outcome information/detail where it could go
<ajs6f>escowles: Okay, but let's not let our impl constraints foul the design of the spec.12:25
<barmintor>what would Heraclitis say about the identity claim of a checksum validation?
just think about it.
<escowles>ajs6f: agreed: the fixity events should look & behave like other events, regardless of what hoops we have to jump through for the mode impl
<ajs6f>escowles: Right.
barmintor: What about Whitehead?12:26
<barmintor>Colson Whitehead?
<ajs6f>I do not know who that is. I was thinking of A.N. Whitehead, the logican/philosopher.12:27
<barmintor>let me get back to you then.
<escowles>whatever we do, let's not use blank nodes or we'll all wind up pondering empty names and reading wittgenstein and russell12:28
<ajs6f>barmintor: Some people who take process philosophy seriously would say that that is literally all you can ever do.
* jcoyne leaves
<ajs6f>escowles: Let's not use blank nodes because THEY ARE A PAIN. Did you ever meet someone who tells a story without anyone's name? Just "What's his face" and "That other woman, what's her name"? That's what working with blank nodes is like.12:29
* escowles no longer has an employer who wants blank node support
<ajs6f>escowles: That's got to feel like putting on a pair of socks filled with freedom.
<barmintor>escowles no longer works for an employer who already wants blank node support12:31
* barmintor sings Time is on my Side
<escowles>barmintor: my current employer does not seem interested in complicated metadata at all12:32
<ajs6f>barmintor: There is a door to freedom, if only you can find it.
<barmintor>Now you always say that you want to be free12:33
But you'll come running back, you'll come running back
You'll come running back to [blank nodes]
<ajs6f>But you will not be able to find them or retrieve them, because YOU FORGOT TO GIVE THEM NAMES.12:34
<barmintor>in the long view of history none of us have names only effects12:37
glittering blank nodes in the firmament of a named graph12:38
our checksums lost like all human knowledge12:39
<ajs6f>barmintor: I don't know how you live your life, but I do my best to avoid having any effects.
<barmintor>here I got you this http://www.imdb.com/title/tt0289879/12:40
what can terrible Ashton Kutcher movies tell us about repositories?12:41
* barmintor proposed the wrong thing for OR2016, damn
<ajs6f>barmintor: http://keycorner.org/pub/image/comic/tdb/1151ckCOMIC-sfpc104-brain-beaker.png Last one on the page.
* dwilcox leaves13:15
* dwilcox joins13:22
* bseeger joins13:24
* dwilcox leaves13:45
* jcoyne joins13:58
* jcoyne leaves
* jcoyne joins14:03
* dhlamb leaves15:48
* diegopino joins15:52
awoods: sorry for bothering you. What is the reason a fedora 4.4.0 can't run on data created on 4.5?
awoods: just for curiosity. because the other way worked fine.15:53
running the fcrepo-webapp jetty console to develop
<awoods>diegopino: that is a little surprising, actually.15:54
diegopino: what error are you seeing?
<diegopino>awoods: sorry for pre-weekend surprises15:56
awoods: ERROR 17:55:00.032 (InvocationContextInterceptor) ISPN000136: Execution error
org.infinispan.persistence.spi.PersistenceException: org.iq80.leveldb.DBException: IO error: /Users/dpino/Desktop/Development/ISLANDORAWORK/Fedor4onclick/fcrepo4-data/fcrepo.ispn.repo.cache/dataFedoraRepository/000102.sst: Invalid argument
<awoods>diegopino: no worries... the week never ends.
<diegopino>and then a lot more, would flood irc
awoods: it's infinispan15:57
<ajs6f>awoods: Actually, my week just ended. Have a good weekend!
* ajs6f leaves
<diegopino>ajs6f ejected15:58
<awoods>diegopino: probably this: https://github.com/fcrepo4/fcrepo4/commit/e0acd0c69229a47791548aa607fdff35fa942d3a
diegopino: associated ticket: https://jira.duraspace.org/browse/FCREPO-1581
<diegopino>awoods: so 7 is nor backwards compat with 616:01
awoods: same happened from 5 to 6
awoods: thanks, curiosity solved!16:02
<awoods>diegopino: I did not see that fact in the upgrade documentation... but that appears to be the case.
diegopino: would you be interested in doing forward/backward compatibility testing for release-candidates in the future?16:03
<diegopino>awoods: probably something changed at cachestore level...
<whikloj>one of us, one of us
<diegopino>awoods: i'm all interested in anything f4. i have been keeping away from pulling and killing bugs just because islandora takes my weeks16:04
awoods: and i'm very islandora involved
awoods: but i will get more involved during this next months.
<awoods>diegopino: that is awesome... and I really appreciate your and whikloj's and ruebot's connection with Islandora.16:05
<diegopino>awoods: =) and we really appreciate what you and the rest do for F416:06
* jcoyne leaves
<diegopino>awoods: ok, have to keep coding. Have a good weekend awoods++ whikloj++ and hugs for everyone!
* diegopino leaves
<barmintor>awoods: Monday on the calendar, thanks!16:12
<awoods>barmintor: thanks16:13
* jgpawletko_ joins16:15
* jgpawletko leaves16:16
* osmandin leaves16:27
* jgpawletko leaves
* barmintor leaves16:29
* kat3_drx leaves16:59
* kat3_drx joins17:00
* kat3_drx leaves
* mcritchlow leaves17:11
* mohamedar leaves17:36
* peichman leaves17:39
* whikloj leaves17:52
* bseeger leaves17:54
* jcoyne joins18:07
* jcoyne leaves18:45
* jcoyne joins19:06
* jcoyne leaves19:23
* jcoyne joins19:25
* mcritchlow joins19:36
* jcoyne leaves19:46
* mcritchlow leaves19:50
* dhlamb joins20:13
* whikloj joins20:38
* jcoyne joins20:54
* whikloj leaves20:55
* whikloj joins20:58
awoods: ping21:07
* whikloj leaves21:18
* dhlamb leaves22:00